- Email: [email protected]
PO-0932: Preliminary clinical study to evaluate an interactive system to segment OARs in thoracic oncology
ESTRO 35 2016
S451
________________________________________________________________________________
perceptual memory (encoding, retrieval, and reaction time to recognize).
Results: ADC is an inverse measure of cellular density. After PRT, average ADC first increased and then decreased; the peaks of the average ADC were detected at 1.5 m and 12 m after PRT for cohort I and II patients. Further, variations in the ADCs were correlated with the mean doses. This dose dependence had different temporal course between the two cohorts. For cohort I, the dose relationship disappeared 12 m after RT. For cohort II, the dose relationship was the strongest at 12 m after RT. ΔADC/ADC (%/Gy) = [0.16, 0.15, 0, -0.06] and [0.16, 0.19, 0.37 0.09] at 1.5, 6, 12 and 24 m after PRT for cohort I and I.FA is a measure of neural connectivity in the brain. On average, no consistent changes in FA were observed for ROIs receiving a mean dose < 40 Gy. In ROIs that received > 40 Gy mean dose, FA decreased consistently. The largest reduction of FA was observed at 1.5 m following PRT. <ΔFA/FA>(%) = [-7.5, -5.3, 2.9, -1.4] at 1.5, 6, 12 and 24 m after PRT for both cohorts. Among NC tests, only changes in verbal and visual semantic retrieval were significant. Decline occurred 1.5 m after PRT (visual semantic reaction time: p<0.005; verbal semantic retrieval: p<000). Recovery occurred 6 m after PRT, and reached baseline at 24 m. Conclusion: ADC and FA are sensitive measures to quantify radiation-induced neuronal injury following PRT. Both ADC and FA showed changes at 1.5 m and a recovery similar to the time course of changes in NC functions. Poster: Physics track: Images and analyses Conclusion: The new method based on local structures in 18F-FDG PET images was a feasible approach. This method is more sensitive in terms of providing a clearer 18FDG uptake dose response six weeks after initiation of treatment compared to standard image subtraction, and may be valuable in future studies addressing RILT. PO-0931 Onset and recovery of neuronal injury following proton radiotherapy C.L. Teng1, M. Mix1, B.K.K. Kevin1, C. Ainsley1, W. Sumei2, K. Manoj2, H. Poptani2, R. Wolf2, L. Sloan1, T. Brown1, N. Thorne1, S. Avery1, Z. Tochner1, C. Hill-Keyser1, S. Mohan2, T. Solberg1, C. Armstrong3, M. Alonson-Basanta1 1 University of Pennsylvania, Radiation Oncology, Philadelphia, USA 2 University of Pennsylvania, Radiology, Philadelphia, USA 3 The Children's Hospital of Philadelphia, Neuron-Oncology, Philadelphia, USA Purpose or Objective: To quantify the time course and the extent of radiation-induced neuronal changes following skull base (cohort I) or brain (cohort II) proton radiation therapy (PRT). Material and Methods: We analyzed 4 cohort I and 4 cohort II patients, who completed 2 year follow-up magnetic resonance imaging (MRI) and neurocognitive (NC) study. Apparent diffusion coefficient (ADC) and fractional anisotropy (FA) from diffusion tensor imaging were used to evaluate neuronal and white matter injury at 1.5, 6, 12 and 24 m following PRT. All MR images for each patient were coregistered to the planning CT using rigid image registration, enabling patient-specific contours (ROIs) to be transferred. Each ROI thus encoded time-dependent MR parameters. The biologically effective doses to GTV ranged from 52 to 70 Gy. Dose-related neuronal changes were compared between the two cohorts as well as within each patient. Cohort I typically received a left-right symmetric PRT with higher dose to the temporal lobes and brainstem, and cohort II a unilateral PRT with a significant higher dose to only hemisphere. ROIs were hippocampus, cerebellum, corpus callosum, temporal lobes, GTV, brainstem and the whole brain. NC testing used 8 memory indices that are radiation-sensitive and insensitive, based on prior series of studies: visual or verbal, semantic or
PO-0932 Preliminary clinical study to evaluate an interactive system to segment OARs in thoracic oncology J. Dolz1,2, H.A. Kirisli1, T. Fechter3, S. Karnitzki3, U. Nestle3, M. Vermandel2, L. Massoptier1 1 AQUILAB Parc Eurasante Lille Metropole, Research, Loos, France 2 Univ. Lille, Inserm- CHU Lille- U1189 - ONCO-THAI - Image Assisted Laser Therapy for Oncology, Lille, France 3 University Medical Center of Freigburg, Department of Radiation Oncology, Freigburg, Germany Purpose or Objective: Radiotherapy aims at delivering the highest possible dose to the tumor while minimizing the irradiation of surrounding healthy tissue, and especially to the organs at risk (OARs). Therefore, accurate delineation of OARs is required for radiation treatment planning (RTP). In thoracic oncology, delineation of some OARs remains manual, making the task time consuming and prone to inter observer variability. Various (semi-) automatic approaches have been proposed to segment OARs on CT but the task still remains challenging. Here, a system to interactively segment OARs in thoracic oncology on CT images is presented and its clinical acceptability evaluated. Material and Methods: The proposed framework has been implemented using MITK platform. User interaction lies in the easy definition of few manual seeds for the OARs and background using a 'paintbrush' tool, which can be interactively added in any view (axial, sagittal or coronal), and is subsequently propagated within the whole volume. Once the user is content with the seeds placement, the system automatically performs the segmentation. If the outcome is not satisfying, the user can modify the seeds, which involves adding and/or removing existing seeds, and perform again the automatic segmentation. Number of tries has been limited to five in the current study. If after the five modifications the segmentation result is not sufficient to be usable in the RTP, the user shall reject it; otherwise, he shall accept it. A hybrid approach combining watershed transformation and graph cuts is used for the segmentation task.
S452 ESTRO 35 2016 ______________________________________________________________________________________________________
Results: The system was evaluated on multivendor CT datasets of 10 patients presenting from early stage to locally advanced NSCLC or pulmonary metastases. OARs taken into consideration in this study were: heart, lungs, oesophagus, proximal bronchus tree, spinal canal and trachea. Interactive contours were generated by a physician using the proposed system. Delineation of the OARs obtained with the presented system was approved to be usable for RTP in more than 90% of the cases, excluding the oesophagus, which segmentation was never approved (Fig 1). On the accepted reported cases, more than 90% of the interactive contours reached a Dice Similarity Coefficient higher than 0.7 with respect to manual segmentations (Fig 2). Therefore, our interactive delineation approach allows users to generate contours of sufficient quality to be used in RTP up to three times faster than manually.
evaluate the accuracy of a single segmentation model trained for optimal segmentation on a variety of different tumour corresponding to different anatomical sites. Material and Methods: ATLAAS, an Automatic decision-Tree based Learning Algorithm for Advanced Segmentation was developed and validated in previous work. ATLAAS (patent pending PCT/GB2015/052981) is a predictive segmentation model, trained with machine learning to automatically select and apply the best PET-AS method, according to the tumour characteristics. The ATLAAS model was trained on 500 simulated PET images with known true contour. The PET-AS used in the model included adaptive iterative thresholding, region growing, watershed-based segmentation, deformable contours, and clustering with K-means, fuzzy C-means and Gaussian Mixture Models, applied to the detection of 2 to 8 clusters. In this work, ATLAAS was applied to the segmentation of PET images containing synthetic tumours generated using the fast PETSTEP simulator. The data included 5 Head and Neck (H&N), 5 lung, 5 abdominal and 5 brain tumours. The contours obtained with ATLAAS were compared to the true tumour outline using the Dice Similarity Coefficient (DSC). DSC results for ATLAAS were compared with results obtained for thresholding at 42% (RT42) and 50% (RT50) of the maximum intensity. Results: ATLAAS contours were closer to the true contour for all cases. The DSCs obtained with ATLAAS were 5.3% to 123% higher across cases than DSCs obtained for RT50 and 4.1% to 74% higher than for RT42. The largest differences between ATLAAS and relative thresholding were obtained for lung images, the smallest differences for H&N and Brain tumours. The minimum conformity of ATLAAS contours on the whole dataset was 0.81 DSC compared to 0.38 and 0.47 for RT50 and RT42 respectively.
Conclusion: An interactive, accurate and easy-to-use computer-assisted system for OARs segmentation in thoracic oncology was presented and clinically evaluated. The introduction of the proposed approach in clinical routine might offer a valuable new option to radiation therapists (RTTs) in performing OARs delineation task. Consequently, further experiments will be carried out on larger databases and with the participation of additional RTTs to investigate its potential use in daily clinical practice. PO-0933 Towards standardisation of PET auto-segmentation with the ATLAAS machine learning algorithm B. Berthon1, C. Marshall2, E. Spezi3 1 Cardiff University, Wales Research and Diagnostic PET Imaging Centre, Cardiff, United Kingdom 2 Cardiff University, Wales Research & Diagnostic PET Imaging Centre, Cardiff, United Kingdom 3 Cardiff University, School of Engineering, Cardiff, United Kingdom Purpose or Objective: Positron Emission Tomography (PET)based -auto-segmentation (PET-AS) methods have been recommended for accurate and reproducible delineation of tumours. However, there is currently no consensus on the best method to use, as different methods have shown better accuracy for different tumour types. This work aims to
Conclusion: Our results show that ATLAAS is capable of providing highly accurate segmentation for different tumour sites, largely outperforming single-value thresholding methods. The ATLAAS machine learning algorithm represents a standardized approach to PET auto-segmentation. The robustness and adaptability of ATLAAS makes it a very promising tool for PET segmentation in radiotherapy treatment planning. PO-0934 Cardio-respiratory motion compensation for 5D thoracic CBCT in IGRT S. Sauppe1, A. Hahn1, M. Brehm2, P. Paysan2, D. Seghers2, M. Kachelrieß1 1 German Cancer Research Center, Medical Physics in Radiology, Heidelberg, Germany 2 Varian Medical Systems, Imaging Laboratory, Baden-Dättwil, Switzerland Purpose or Objective: Accurate information about patient motion is essential for precise radiation therapy, in particular for thoracic and abdominal cases. Patient motion assessment based on daily on-board CBCT images immediately before treatment potentially allows accounting for organ motion during the treatment. Especially for patients with tumors close to organs at risk, the organ positions need to be precisely known as a function of time. In case of the heart 5D
S451
________________________________________________________________________________
perceptual memory (encoding, retrieval, and reaction time to recognize).
Results: ADC is an inverse measure of cellular density. After PRT, average ADC first increased and then decreased; the peaks of the average ADC were detected at 1.5 m and 12 m after PRT for cohort I and II patients. Further, variations in the ADCs were correlated with the mean doses. This dose dependence had different temporal course between the two cohorts. For cohort I, the dose relationship disappeared 12 m after RT. For cohort II, the dose relationship was the strongest at 12 m after RT. ΔADC/ADC (%/Gy) = [0.16, 0.15, 0, -0.06] and [0.16, 0.19, 0.37 0.09] at 1.5, 6, 12 and 24 m after PRT for cohort I and I.FA is a measure of neural connectivity in the brain. On average, no consistent changes in FA were observed for ROIs receiving a mean dose < 40 Gy. In ROIs that received > 40 Gy mean dose, FA decreased consistently. The largest reduction of FA was observed at 1.5 m following PRT. <ΔFA/FA>(%) = [-7.5, -5.3, 2.9, -1.4] at 1.5, 6, 12 and 24 m after PRT for both cohorts. Among NC tests, only changes in verbal and visual semantic retrieval were significant. Decline occurred 1.5 m after PRT (visual semantic reaction time: p<0.005; verbal semantic retrieval: p<000). Recovery occurred 6 m after PRT, and reached baseline at 24 m. Conclusion: ADC and FA are sensitive measures to quantify radiation-induced neuronal injury following PRT. Both ADC and FA showed changes at 1.5 m and a recovery similar to the time course of changes in NC functions. Poster: Physics track: Images and analyses Conclusion: The new method based on local structures in 18F-FDG PET images was a feasible approach. This method is more sensitive in terms of providing a clearer 18FDG uptake dose response six weeks after initiation of treatment compared to standard image subtraction, and may be valuable in future studies addressing RILT. PO-0931 Onset and recovery of neuronal injury following proton radiotherapy C.L. Teng1, M. Mix1, B.K.K. Kevin1, C. Ainsley1, W. Sumei2, K. Manoj2, H. Poptani2, R. Wolf2, L. Sloan1, T. Brown1, N. Thorne1, S. Avery1, Z. Tochner1, C. Hill-Keyser1, S. Mohan2, T. Solberg1, C. Armstrong3, M. Alonson-Basanta1 1 University of Pennsylvania, Radiation Oncology, Philadelphia, USA 2 University of Pennsylvania, Radiology, Philadelphia, USA 3 The Children's Hospital of Philadelphia, Neuron-Oncology, Philadelphia, USA Purpose or Objective: To quantify the time course and the extent of radiation-induced neuronal changes following skull base (cohort I) or brain (cohort II) proton radiation therapy (PRT). Material and Methods: We analyzed 4 cohort I and 4 cohort II patients, who completed 2 year follow-up magnetic resonance imaging (MRI) and neurocognitive (NC) study. Apparent diffusion coefficient (ADC) and fractional anisotropy (FA) from diffusion tensor imaging were used to evaluate neuronal and white matter injury at 1.5, 6, 12 and 24 m following PRT. All MR images for each patient were coregistered to the planning CT using rigid image registration, enabling patient-specific contours (ROIs) to be transferred. Each ROI thus encoded time-dependent MR parameters. The biologically effective doses to GTV ranged from 52 to 70 Gy. Dose-related neuronal changes were compared between the two cohorts as well as within each patient. Cohort I typically received a left-right symmetric PRT with higher dose to the temporal lobes and brainstem, and cohort II a unilateral PRT with a significant higher dose to only hemisphere. ROIs were hippocampus, cerebellum, corpus callosum, temporal lobes, GTV, brainstem and the whole brain. NC testing used 8 memory indices that are radiation-sensitive and insensitive, based on prior series of studies: visual or verbal, semantic or
PO-0932 Preliminary clinical study to evaluate an interactive system to segment OARs in thoracic oncology J. Dolz1,2, H.A. Kirisli1, T. Fechter3, S. Karnitzki3, U. Nestle3, M. Vermandel2, L. Massoptier1 1 AQUILAB Parc Eurasante Lille Metropole, Research, Loos, France 2 Univ. Lille, Inserm- CHU Lille- U1189 - ONCO-THAI - Image Assisted Laser Therapy for Oncology, Lille, France 3 University Medical Center of Freigburg, Department of Radiation Oncology, Freigburg, Germany Purpose or Objective: Radiotherapy aims at delivering the highest possible dose to the tumor while minimizing the irradiation of surrounding healthy tissue, and especially to the organs at risk (OARs). Therefore, accurate delineation of OARs is required for radiation treatment planning (RTP). In thoracic oncology, delineation of some OARs remains manual, making the task time consuming and prone to inter observer variability. Various (semi-) automatic approaches have been proposed to segment OARs on CT but the task still remains challenging. Here, a system to interactively segment OARs in thoracic oncology on CT images is presented and its clinical acceptability evaluated. Material and Methods: The proposed framework has been implemented using MITK platform. User interaction lies in the easy definition of few manual seeds for the OARs and background using a 'paintbrush' tool, which can be interactively added in any view (axial, sagittal or coronal), and is subsequently propagated within the whole volume. Once the user is content with the seeds placement, the system automatically performs the segmentation. If the outcome is not satisfying, the user can modify the seeds, which involves adding and/or removing existing seeds, and perform again the automatic segmentation. Number of tries has been limited to five in the current study. If after the five modifications the segmentation result is not sufficient to be usable in the RTP, the user shall reject it; otherwise, he shall accept it. A hybrid approach combining watershed transformation and graph cuts is used for the segmentation task.
S452 ESTRO 35 2016 ______________________________________________________________________________________________________
Results: The system was evaluated on multivendor CT datasets of 10 patients presenting from early stage to locally advanced NSCLC or pulmonary metastases. OARs taken into consideration in this study were: heart, lungs, oesophagus, proximal bronchus tree, spinal canal and trachea. Interactive contours were generated by a physician using the proposed system. Delineation of the OARs obtained with the presented system was approved to be usable for RTP in more than 90% of the cases, excluding the oesophagus, which segmentation was never approved (Fig 1). On the accepted reported cases, more than 90% of the interactive contours reached a Dice Similarity Coefficient higher than 0.7 with respect to manual segmentations (Fig 2). Therefore, our interactive delineation approach allows users to generate contours of sufficient quality to be used in RTP up to three times faster than manually.
evaluate the accuracy of a single segmentation model trained for optimal segmentation on a variety of different tumour corresponding to different anatomical sites. Material and Methods: ATLAAS, an Automatic decision-Tree based Learning Algorithm for Advanced Segmentation was developed and validated in previous work. ATLAAS (patent pending PCT/GB2015/052981) is a predictive segmentation model, trained with machine learning to automatically select and apply the best PET-AS method, according to the tumour characteristics. The ATLAAS model was trained on 500 simulated PET images with known true contour. The PET-AS used in the model included adaptive iterative thresholding, region growing, watershed-based segmentation, deformable contours, and clustering with K-means, fuzzy C-means and Gaussian Mixture Models, applied to the detection of 2 to 8 clusters. In this work, ATLAAS was applied to the segmentation of PET images containing synthetic tumours generated using the fast PETSTEP simulator. The data included 5 Head and Neck (H&N), 5 lung, 5 abdominal and 5 brain tumours. The contours obtained with ATLAAS were compared to the true tumour outline using the Dice Similarity Coefficient (DSC). DSC results for ATLAAS were compared with results obtained for thresholding at 42% (RT42) and 50% (RT50) of the maximum intensity. Results: ATLAAS contours were closer to the true contour for all cases. The DSCs obtained with ATLAAS were 5.3% to 123% higher across cases than DSCs obtained for RT50 and 4.1% to 74% higher than for RT42. The largest differences between ATLAAS and relative thresholding were obtained for lung images, the smallest differences for H&N and Brain tumours. The minimum conformity of ATLAAS contours on the whole dataset was 0.81 DSC compared to 0.38 and 0.47 for RT50 and RT42 respectively.
Conclusion: An interactive, accurate and easy-to-use computer-assisted system for OARs segmentation in thoracic oncology was presented and clinically evaluated. The introduction of the proposed approach in clinical routine might offer a valuable new option to radiation therapists (RTTs) in performing OARs delineation task. Consequently, further experiments will be carried out on larger databases and with the participation of additional RTTs to investigate its potential use in daily clinical practice. PO-0933 Towards standardisation of PET auto-segmentation with the ATLAAS machine learning algorithm B. Berthon1, C. Marshall2, E. Spezi3 1 Cardiff University, Wales Research and Diagnostic PET Imaging Centre, Cardiff, United Kingdom 2 Cardiff University, Wales Research & Diagnostic PET Imaging Centre, Cardiff, United Kingdom 3 Cardiff University, School of Engineering, Cardiff, United Kingdom Purpose or Objective: Positron Emission Tomography (PET)based -auto-segmentation (PET-AS) methods have been recommended for accurate and reproducible delineation of tumours. However, there is currently no consensus on the best method to use, as different methods have shown better accuracy for different tumour types. This work aims to
Conclusion: Our results show that ATLAAS is capable of providing highly accurate segmentation for different tumour sites, largely outperforming single-value thresholding methods. The ATLAAS machine learning algorithm represents a standardized approach to PET auto-segmentation. The robustness and adaptability of ATLAAS makes it a very promising tool for PET segmentation in radiotherapy treatment planning. PO-0934 Cardio-respiratory motion compensation for 5D thoracic CBCT in IGRT S. Sauppe1, A. Hahn1, M. Brehm2, P. Paysan2, D. Seghers2, M. Kachelrieß1 1 German Cancer Research Center, Medical Physics in Radiology, Heidelberg, Germany 2 Varian Medical Systems, Imaging Laboratory, Baden-Dättwil, Switzerland Purpose or Objective: Accurate information about patient motion is essential for precise radiation therapy, in particular for thoracic and abdominal cases. Patient motion assessment based on daily on-board CBCT images immediately before treatment potentially allows accounting for organ motion during the treatment. Especially for patients with tumors close to organs at risk, the organ positions need to be precisely known as a function of time. In case of the heart 5D